The agency of the critical discourse analyst in relation to ChatGPT
a comparative study based on the #AstesianoCase in Uruguay
DOI:
https://doi.org/10.1590/1983-3652.2026.63166Keywords:
Generative Artificial Intelligence, Critical Discourse Analysis, Investigative agency, Political corruption, Appraisal frameworkAbstract
This article examines the agency of the critical discourse analyst in the context of the incorporation of Generative Artificial Intelligence (GAI), through an empirical comparison between the analysis conducted by human researchers and the processing performed by ChatGPT on the same discourse corpus related to the #AstesianoCase in Uruguay. The analysis was carried out in four phases: corpus construction, critical discourse analysis conducted by the researchers, replication of the same stages using ChatGPT Plus 5.1, and comparison between both sets of results. A corpus of 431 tweets mentioning the account of then-president @luislacallepou was compiled from an initial sample of 20,000 tweets obtained through scraping and processing in R, with methodological support from ChatGPT Plus. The human analysis classified the texts according to the support, criticism, or indeterminacy expressed by users toward the former president, identified three thematic axes of criticism, and applied the Appraisal framework (Martin; White, 2005). ChatGPT was instructed to replicate the analytical operations. The comparative analysis showed that the tool demonstrated utility in technical tasks; however, it revealed interpretative limitations, including errors in the classification of users’ positioning, partial identification of thematic axes, and restricted application of the Appraisal framework. Under the specific conditions of this study and for the types of interpretative operations examined, the limitations of automated analysis in comparison with human discourse analysis ultimately reinforce the need for situated interpretation in critical discourse research.
Downloads
References
APARICI, Roberto y GARCÍA-MARTÍN, David. Prosumidores y emirecs: análisis de dos teorías enfrentadas. Comunicar, volumen 26, número 55, páginas 71-79, 2018. DOI: 10.3916/C55-2018-07.
BUCKINGHAM, David. Educación en medios: alfabetización, aprendizaje y cultura contemporánea. Buenos Aires: Paidós, 2005.
CURRY, Niall; BAKER, Paul y BROOKES, Gavin. Generative AI for Corpus Approaches to Discourse Studies: A Critical Evaluation of ChatGPT. Applied Corpus Linguistics, volumen 4, número 1, 2024. artículo 100082. DOI: 10.1016/j.acorp.2023.100082.
DIJK, Teun A. van. Política, ideología y discurso. En: BROWN, E. K. (editor). Encyclopedia of Language and Linguistics. Oxford: Elsevier, 2006. páginas 728-740.
DIJK, Teun A. van. Discourse and Power. London: Palgrave Macmillan, 2008.
FAIRCLOUGH, Norman. Critical Discourse Analysis: The Critical Study of Language. London: Longman, 1995.
FONT-JULIÁN, Cristina; ORDUÑA-MALEA, Enrique y CODINA, Lluís. ChatGPT Search as a Tool for Scholarly Tasks: Evolution or Devolution? Infonomy, volumen 2, número 5, 2024. DOI: 10.3145/infonomy.24.077. Disponible en: https://infonomy.scimagoepi.com/index.php/infonomy/article/view/77.
GAMBETTA, Leticia. El (Des)equilibrio Discursivo en la Red: Una Propuesta de Categorización de Enunciadores. En: IV Congreso Internacional sobre Movimientos Sociales y TICs. Sevilla: [s. n.], 2021.
GAMBETTA, Leticia; GONZÁLEZ, Manuel y CASTRO, Leticia. Narrativas periodísticas y disputa de sentidos sobre la corrupción: el caso Astesiano en Uruguay. Discurso & Sociedad, volumen 21, número 1, 2027. No prelo.
GAMBETTA, Leticia y HUERTAS, Agustina. El enunciador común en las redes sociales: la eutanasia en debate. Comunicação Midiática, volumen 17, número 2, páginas 139-156, 2022. Disponible en: https://www2.faac.unesp.br/comunicacaomidiatica/index.php/CM/article/view/540.
GAMBETTA ABELLA, Leticia y HUERTAS, Agustina. Media That No Longer Mediate: The Blurred Enunciation in Contemporary Journalistic Practices. Brazilian Journalism Research, volumen 20, número 3, e1683, 2024. DOI: 10.25200/BJR.v20n3.2024.1683.
GIL DE ZÚÑIGA, Homero; GOYANES, Manuel y DUROTOYE, Timilehin. A Scholarly Definition of Artificial Intelligence (AI): Advancing AI as a Conceptual Framework in Communication Research. Political Communication, volumen 41, número 2, páginas 317-334, 2024. DOI: 10.1080/10584609.2023.2290493.
GILLINGS, Mathew; KOHN, Tobias y MAUTNER, Gerlinde. The Rise of Large Language Models: Challenges for Critical Discourse Studies. Critical Discourse Studies, páginas 625-641, 2024. DOI: 10.1080/17405904.2024.2373731.
HALLIDAY, M. A. K. y MATTHIESSEN, Christian M. I. M. An Introduction to Functional Grammar. 3. edición. London: Hodder Education, 2004.
KHOSRAVINIK, Majid. Social Media Critical Discourse Studies (SM-CDS). En: FLOWERDEW, John y RICHARDSON, John E. (editores). The Routledge Handbook of Critical Discourse Studies. London: Routledge, 2017. páginas 582-596. DOI: 10.4324/9781315739342-40.
KHOSRAVINIK, Majid. Digital Meaning-Making across Content and Practice in Social Media Critical Discourse Studies. Critical Discourse Studies, volumen 19, número 2, páginas 119-123, 2022. DOI: 10.1080/17405904.2020.1835683.
LA DIARIA. Jefe de la custodia presidencial tiene más de veinte indagatorias policiales por múltiples delitos. Sept. 2021. Disponible en: https://ladiaria.com.uy/articulo/2021/9/jefe-de-la-custodia-presidencial-tiene-mas-de-veinte-indagatorias-policiales-por-multiples-delitos/. Acceso en: 10 jul. 2025.
LA DIARIA. Fue detenido el jefe de custodia de Lacalle Pou por falsificar documentación para la expedición de pasaportes a ciudadanos rusos. Sept. 2022. Disponible en: https://ladiaria.com.uy/justicia/articulo/2022/9/fue-detenido-el-jefe-de-custodia-de-lacalle-pou-por-falsificar-documentacion-para-la-expedicion-de-pasaportes-a-ciudadanos-rusos/. Acceso en: 15 agosto 2025.
MACI, Stefania y ANESA, Patrizia. The Impact of AI on Discourse Analysis: Challenges and Opportunities. International Journal of Language Studies, volumen 19, número 2, páginas 1-12, 2025.
MARTIN, J. R. y WHITE, P. R. R. The Language of Evaluation: Appraisal in English. New York: Palgrave Macmillan, 2005.
MARTÍNEZ-FRESNEDA OSORIO, Humberto y SÁNCHEZ RODRÍGUEZ, Gabriel. La influencia de Twitter en la agenda setting de los medios de comunicación. Revista de Ciencias de la Comunicación e Información, volumen 27, 2022. DOI: 10.35742/rcci.2022.27.e136.
MÜLLER, Pia. Estudios críticos del discurso y discursos en los medios sociales: retos teóricos y metodológicos y posibles soluciones. Journal of Discourse Research, número 1, páginas 19-40, 2020.
MVD NOTICIAS. #MVDNoticias A través de chat al que accedió #MVDNoticias [...] X. Dic. 2022. Disponible en: https://twitter.com/MVDNoticias/status/1605325090765283330.
NOBLÍA, Adriana. La política en tiempos de redes sociales: discursos, identidades y conflictos. Buenos Aires: Teseo, 2020.
OTEÍZA, Teresa y PINUER, Claudio. El sistema de valoración como herramienta teórico-metodológica para el estudio social e ideológico del discurso. Logos, volumen 29, número 2, 2019. DOI: 10.15443/RL2918.
TING, Liang; SINGH, Charanjit K. S.; KIONG, Tee Tze; RAKHIMJONOV, Farkhodjon y SINGH, Tarsame S. M. Mentor or Examiner? A Critical Discourse Analysis of AI-Generated Feedback in EFL Writing Education. International Journal of Academic Research in Progressive Education and Development, volumen 14, número 4, páginas 664-684, 2025. DOI: 10.6007/IJARPED/v14-i4/26651.
WODAK, Ruth y MEYER, Michael. Métodos de Análise Crítica do Discurso. São Paulo: Gedisa, 2003.
ZAPPAVIGNA, Michele. The Discourse of Twitter and Social Media: How We Use Language to Create Affiliation on the Web. London: Continuum, 2012.
ZAPPAVIGNA, Michele. Hack Your Corpus Analysis: How AI Can Assist Corpus Linguists Deal with Messy Social Media Data. Applied Corpus Linguistics, volumen 3, página 100071, 2023. DOI: 10.1016/j.acorp.2023.100071.
Downloads
Published
Data Availability Statement
Los datos de investigación solo están disponibles previa solicitud.
Issue
Section
License
Copyright (c) 2026 Leticia Gambetta Abella, Andrés Abdala, Agustina Huertas Rehermann

This work is licensed under a Creative Commons Attribution 4.0 International License.
This is an open access article that allows unrestricted use, distribution and reproduction in any medium as long as the original article is properly cited.








